A triangulated \(K3\) surface with the minimum number of vertices (Q5950612)

The authors give an explicit description of a 16-vertex triangulation of a K3 surface. It follows from results by \textit{W. Kühnel} [Tight polyhedral submanifolds and tight triangulations, Lect. Notes Math. 1612 (1995; Zbl 0834.53004)] that there is no triangulation of a K3 surface with fewer vertices. The triangulation was obtained by an exhaustive computer search for all \(4\)-dimensional pseudo-manifolds on \(16\)~vertices which admit a primitive (in particular transitive) action on the vertex set. It turns out that except for the K3 surface only a previously known triangulated Kummer variety with \(16\) singularities occurs. The K3 surface is identified as follows: It is verified that each vertex link is a triangulated sphere, hence the pseudo-manifold is a manifold. Each triplet of vertices is a face, hence the manifold is simply connected. From an explicit computation it is deduced that the second Stiefel-Whitney class vanishes. Therefore the intersection form is even. Finally, the signature is determined to be \(\pm 16\). The result now follows from \textit{M. H. Freedman}'s characterization [J. Differ. Geom. 17, 357-453 (1982; Zbl 0528.57011)] of simply connected \(4\)-manifolds.